ChristopherSchultz/fast-file-count

Branch: master

fast-file-count / dircnt.c

1	/**
2	* dircnt.c - a fast file-counting program.
3	*
4	* Written 2015-02-06 by Christopher Schultz as a programming demonstration
5	* for a StackOverflow answer:
6	* https://stackoverflow.com/questions/1427032/fast-linux-file-count-for-a-large-number-of-files/28368788#28368788
7	*
8	* This code is licensed under the Apache License 2.0. Please read the file
9	* LICENSE for more information.
10	*
11	* Please see the README.md file for compilation and usage instructions.
12	*
13	* Thanks to FlyingCodeMonkey, Gary R. Van Sickle, and Jonathan Leffler for
14	* various suggestions and improvements to the original code. Any additional
15	* contributors can be found by looking at the GitHub revision history from
16	* this point forward..
17	*/
18	#include <stdio.h>
19	#include <dirent.h>
20	#include <string.h>
21	#include <stdlib.h>
22	#include <limits.h>
23	#include <sys/stat.h>
24
25	#if defined(WIN32) \|\| defined(_WIN32)
26	#define PATH_SEPARATOR '\\'
27	#else
28	#define PATH_SEPARATOR '/'
29	#endif
30
31	#define EXIT_REACHED_LIMIT 0x01
32
33	/* A custom structure to hold separate file and directory counts */
34	struct filecount {
35	unsigned long dirs;
36	unsigned long files;
37	};
38
39	/*
40	* counts the number of files and directories in the specified directory.
41	*
42	* path - relative pathname of a directory whose files should be counted
43	* counts - pointer to struct containing file/dir counts
44	*/
45	void count(char path, struct filecount counts) {
46	DIR dir; / dir structure we are reading */
47	struct dirent ent; / directory entry currently being processed */
48	char subpath[PATH_MAX]; /* buffer for building complete subdir and file names */
49	struct stat statbuf; /* buffer for stat() info. A call to lstat() might be
50	required even if _DIRENT_HAVE_D_TYPE is true
51	because ent->d_type might be DT_UNKNOWN */
52	int isdir; /* flag for a directory entry being a directory */
53
54	#ifdef DEBUG
55	fprintf(stderr, "Opening dir %s\n", path);
56	#endif
57	dir = opendir(path);
58
59	/* opendir failed... file likely doesn't exist or isn't a directory */
60	if(NULL == dir) {
61	perror(path);
62	return;
63	}
64
65	while((ent = readdir(dir))) {
66	if (strlen(path) + 1 + strlen(ent->d_name) > PATH_MAX) {
67	fprintf(stdout, "path too long (%ld) %s%c%s", (strlen(path) + 1 + strlen(ent->d_name)), path, PATH_SEPARATOR, ent->d_name);
68	return;
69	}
70
71	isdir = 0; /* reset isdir flag */
72	#ifdef DEBUG
73	fprintf(stderr, "Considering %s%c%s\n", path, PATH_SEPARATOR, ent->d_name);
74	#endif /* DEBUG */
75
76	/* Use dirent.d_type if present, otherwise use stat() */
77	#if ( defined ( _DIRENT_HAVE_D_TYPE ) && !PREFER_STAT)
78	if(DT_UNKNOWN == ent->d_type) {
79	/* Must perform lstat() anyway */
80	#ifdef DEBUG
81	fprintf(stderr, "Dirent type is DT_UNKNOWN, must perform lstat()\n");
82	#endif /* DEBUG */
83	sprintf(subpath, "%s%c%s", path, PATH_SEPARATOR, ent->d_name);
84	if(lstat(subpath, &statbuf)) {
85	perror(subpath);
86	return;
87	}
88	if(S_ISDIR(statbuf.st_mode)) {
89	#ifdef DEBUG
90	fprintf(stderr, "Determined %s is a directory via lstat(1)\n", subpath);
91	#endif /* DEBUG */
92	isdir = 1;
93	}
94	} else if(DT_DIR == ent->d_type) {
95	#ifdef DEBUG
96	fprintf(stderr, "Determined %s%c%s is a directory via dirent\n", path, PATH_SEPARATOR, ent->d_name);
97	#endif /* DEBUG */
98	isdir = 1;
99	}
100	#else
101	sprintf(subpath, "%s%c%s", path, PATH_SEPARATOR, ent->d_name);
102	if(lstat(subpath, &statbuf)) {
103	perror(subpath);
104	return;
105	}
106
107	if(S_ISDIR(statbuf.st_mode)) {
108	#ifdef DEBUG
109	fprintf(stderr, "S_ISDIR=%d, mode bits=%x\n", S_ISDIR(statbuf.st_mode), statbuf.st_mode);
110	fprintf(stderr, "Determined %s is a directory via lstat(2)\n", subpath);
111	#endif /* DEBUG */
112	isdir = 1;
113	}
114	#endif /* if defined _DIRENT_HAVE_D_TYPE, etc. */
115
116	#ifdef DEBUG
117	fprintf(stderr, "name=%s, isdir=%d\n", ent->d_name, isdir);
118	#endif
119
120	if(isdir) {
121	/* Skip "." and ".." directory entries... they are not "real" directories */
122	if(0 == strcmp("..", ent->d_name) \|\| 0 == strcmp(".", ent->d_name)) {
123	/* fprintf(stderr, "This is %s, skipping\n", ent->d_name); */
124	} else {
125	if(ULONG_MAX == counts->dirs) {
126	fprintf(stderr, "Reached maximum number of directories to count (%lu) after %lu files\n", counts->dirs, counts->files);
127	exit(EXIT_REACHED_LIMIT);
128	}
129	sprintf(subpath, "%s%c%s", path, PATH_SEPARATOR, ent->d_name);
130	counts->dirs++;
131	count(subpath, counts);
132	}
133	} else {
134	if(ULONG_MAX == counts->files) {
135	fprintf(stderr, "Reached maximum number of files to count (%lu) after %lu directories\n", counts->files, counts->dirs);
136	exit(EXIT_REACHED_LIMIT);
137	}
138
139	counts->files++;
140	}
141	}
142
143	#ifdef DEBUG
144	fprintf(stderr, "Closing dir %s\n", path);
145	#endif
146	closedir(dir);
147	}
148
149	int main(int argc, char *argv[]) {
150	struct filecount counts;
151	char *dir;
152	counts.files = 0;
153	counts.dirs = 0;
154	if(argc > 1)
155	dir = argv[1];
156	else
157	dir = ".";
158
159	#ifdef DEBUG
160	#if PREFER_STAT
161	fprintf(stderr, "Compiled with PREFER_STAT. Using lstat()\n");
162	#elif defined ( _DIRENT_HAVE_D_TYPE )
163	fprintf(stderr, "Using dirent.d_type\n");
164	#else
165	fprintf(stderr, "Don't have dirent.d_type, falling back to using lstat()\n");
166	#endif
167	#endif
168
169	count(dir, &counts);
170
171	/* If we found nothing, this is probably an error which has already been printed */
172	if(0 < counts.files \|\| 0 < counts.dirs) {
173	printf("%s contains %lu files and %lu directories\n", dir, counts.files, counts.dirs);
174	}
175
176	return 0;
177	}
178

Terms
Privacy
Security
Status
Help